Do the accents sound right?

• Apr 4, 2024 - 23:04

My first and most important question is: is it correct that after a sudden accent, the notes continue to sound in the dynamic range that was specified before the accent? Here is a simple video example: first there was a piano, then a crescendo to the sfz accent, and after that the volume drops to the initial piano. Is this correct? 00-57-53_1.ts On the one hand, the accents sfz, sff, fz, etc. act only on one note/chord, and on the other hand, there was a crescendo, but after that the sound became like a piano again.
And here's my second question. Do the accents in Musescore currently correspond to alphabetic characters in the way shown in the table below? Снимок экрана (210).png

Attachment Size
Снимок экрана (210).png 124.57 KB
00-57-53_1.ts 793.13 KB


Well, I am very sorry that no one can give an answer to the questions I have raised. It seems that this will remain a mystery.

Both articulations like accent markings and special dynamic markings like sfz apply to just a single note, then the original dynamic is expected to return afterward.

That table is a nice subjective approximation of one person's opinion of the possible interpretation of these markings in whatever style of music they were referring to. But there is no objectively correct answer here. MuseScore's playback does what the professionals who designed it thought was most appropriate, which may or may not happen to coincide in some cases with what the person who constructed that table was thinking about when they created it.

In reply to by Marc Sabatella

Do you have data on what velocity is used for each accent/verbal indication? (For example, "mf" — 80; "sfz" — 112; ^ together with "ff" — 126; etc.)
In my opinion, this should be written somewhere in the source code of the software, but I can't even imagine how and where to find it.

In reply to by Dima S.

I guess you’re talking about sound fonts, since you’re referencing MIDI velocity numbers? Muse Sounds doesn’t really operate on that same principle, and it’s not open source, so you won’t find that info. But for sound fonts, then indeed the values are probably in there somewhere. That’s all about to change for 4.4 though.

In reply to by Marc Sabatella

Yes, first of all I mean the sound font MuseScore Basic. Although I would also like to find out the same thing about Muse Sounds, but since you just said yourself that Muse Sounds has a closed source code, and therefore you won't be able to find out anything there, but nevertheless, I think that the velocity of Muse Sounds was done using the example of the velocity of MuseScore Basic.

In reply to by Dima S.

Again, Muse Sounds doesn’t use MIDI internally, and was developed by an entirely separate team, so I wouldn’t expect there is any particular relationship to how it interprets dynamics compared to MS Basic. In any case, the velocity settings used for sound fonts aren’t in the sound font itself; they are MuseScore source code and are interpreted further by the “fluidsynth” synthesizer used by MuseScore.

Is there a reason you’re curious about the internal details?

In reply to by Marc Sabatella

I got the reason when I saw that very table. I agree that not everything in this table is unambiguous, and there is no clear rule for determining the interpretation of accents. However, I have never seen such an interpretation anywhere, and no one could answer me by what accent to understand this or that "one-time" dynamic indication. The table became a miracle for me, and I wanted it to become standardized. However, I am interested in how Musescore interprets these instructions. Maybe the "professionals" (quote) interpret it in a completely different way, but I do not know this. And here it should be understood that although this table was created as the opinion of one person, but his opinion should be respected, and not considered as some kind of guess and assumption. This table is taken from the book "Essential Dictionary of Music Notation" by Tom Gerou and Linda Lusk. I do not know if these people are professionals, but at least for the first time in history (correct me if I'm wrong) they created such a table, and nowhere in their book does it say that such a table is "approximate" (in any case, it also does not say that this table is an accurate interpretation, but nevertheless, let's respect what they wrote).
If each professional understands the accented dynamic instructions in his own way, then musical chaos and ambiguity will result. In the end, take a look at the book I mentioned, lay down what is given in it as the basis for software. These designations are in the SMuFL, but not everyone understands the difference between them. There are discussions and disputes about this. I've been reading discussions for a long time about what the difference is, for example, between "sf", "fz" and "sfz". Everyone had their own opinions. Someone claimed that there was no difference at all. But here's a book for you, here's a table for you, here's the answer to a long and seemingly unsolvable question. It's time to sort out these interpretations. I wonder if these professionals have even seen this table? Whether they took it into account or decided that it was bullshit, and we will interpret it all in our own way. That's why I brought up this topic, because it's important for me to understand it.
By the way, let me clarify, you mentioned that you wrote more than a hundred lines of code for such a wonderful software as Musescore. In which area exactly are you a contributor? For example, Simon (oktophonie) is the head of the engraving department, Ben Averill (avvvvve) is responsible for the design. What are you responsible for?

In reply to by Dima S.

It’s not that all “professionals” die out one way and this table is different. Literal;lay person on earth interests dynamics differently. There just isn’t a standard, and no amount of tables found online will cause musicians to suddenly all start interpretations everything the same way. Just as different musicians play “f” at a different volume, or different in one ensemble from another, similar for how they interpret sfz or accent. Chaos doesn’
T result because musicians in an ensemble listen to each other and adjust as necessary. Conductors can also give feedback (“horns - please play measure 57 a little more quietly”). The same same musician playing in three different ensembles might play “mp“ at three different volume levels - and perhaps at different volume levels at different places within the same piece depending on what other instruments he or she is grouped with for each passage.

So it’s not a question of “respecting” what any one particular person decided to write down in a chart for the benefit of beginners who might not otherwise have known where to at least start. It’s just a matter of recognizing that this is all such charts can possibly be - a starting point for developing your own interpretation based on further experience.

Bottom line: understand the code used to implement things under the hood has absolutely nothing to do with becoming a better musician. If you want to learn how to interpret these symbols better, just get more experience playing in ensembles. The numbers from a table or numbers found in the code aren’t going to be useful to you. Even if so ine could tell you absolutely that sfz should be played at a MIDI velocity of 102 (which is of course completely not possible), it would be also impossible for you to maker use of that information in your own playing. Grow hard to you need to blow into a trumpet to reach a MIDI value of 102?

As for my contributions for MuseScore, my first 100 lines fop code were back in 2012 or so :-). By now it’s hundreds of thousands, I suspect. It’s no longer easy to get counts from GitHub because the project has become too big. Most of my code has to do with the engraving side of things and also some editing commands - very little about playback. But almost all of that before the current team came in.

In reply to by Marc Sabatella

You're saying everything right, but this is about playback in real life. I'm talking about the program sound. It's not about irl, it's about playback.
For example, according to the table I have given (I'm lying, not by me, but by the authors, also mentioned by me above), "fz" should be interpreted as an accent in the dynamic range from "ppp" to "f". In Musescore (at the moment) this dynamic designation [fz] does not change the sound at all, because of this, it seems that professionals simply gave up on it, and, as a result, there is doubt about the correct interpretation of other similar dynamic instructions. That's why I'm asking for a list of interpretations of such accents by velocity.
At least in Musescore 3 there were forms in which the velocity and a certain "velocity change" were indicated (by the way, I also ask you to clarify what a "velocity change" is). Снимок экрана (304).png There is no such thing in Musescore 4 anymore (most likely due to the introduction of Muse Sounds).
About the code… It turns out that you are now, so to speak, a "veteran" in terms of contributing in writing code (if I understand correctly that after the arrival of the current team, your contribution became worthless for them). If it's not a secret, what exactly have you implemented in Musescore? What are the main (maybe the most memorable) innovations/fixes that you have made? I apologize for being so deeply curious, but I really wonder what contribution this or that contributor made. In this regard, I am not indifferent.

In reply to by Dima S.

MuseScore 4 will eventually have similar controls over the behavior of accents and sfz, but again, since Muse Sounds isn't based on MIDI, it will almost certainly work differently than the old MuseScore 3 controls.

If you encounter a dynamic marking that does nothing and you believe it should, just open an issue on GitHub. Then the developers who work on sound can check it out. They won't need you to quote tables about what someone on the internet says, they'll just apply their own judgement. Right now I can confirm fz (which is not really a common symbol) just doesn't seem to do anything; sfz is a better choice for playback. But there's no reason they couldn't make fz just work the same as sfz once you open up the request on GitHub.

As for my contributions, It's not that they became worthless, I just stopped having the sort of time needed to keep up with the development. I continue to fix a couple of random issues here and there but mostly I am happy to let others take over. I had some involvement in almost every area of the program, but my most noteworthy contributions I can think of offhand were the chord symbol parser, lots of details about how accidentals, ties, slurs, and other symbols are positioned, the system to allow manual adjustments without the need to disable automatic placement, some editing commands like explode/implode and paste half/double duration, and things having to do with accessibility for blind users. Plenty more that just isn't coming to mind at the moment. I also led the projects to implement swing and chord symbol playback, but didn't actually work on the code myself much - playback is the area I have touched the least.

In reply to by Dima S.

I'm afraid that you are mistaken in several ways. That table is nothing but one person's very subjective opinion on what the symbols mean. It is not standard; it is not correct. (Equally, it is not incorrect!) Every other musician will have a different opinion, varying from slightly to radically different than this one author. Any attempt to standardize this will lead to complaints from those with other opinions that "THAT'S NOT RIGHT".

Second, fz does change the sound. (At least in the MuS Basic sound; I didn't try the MuseSounds.) Fairly small, but try setting the dynamic level very soft (e.g., ppp***) and add a **fz. So, the programmers definitely did not give up on it. They gave it one interpretation ... one that is presumably as "correct" (or "incorrect") as any other.

Last, now is exactly the wrong time to be concerned about this. As Marc said, the sound of playback is in the process of a massive change. You'd be best off to not worry about it now. Wait until the new playback sound system is completed.

In reply to by TheHutch

I just checked and concluded that "fz" is not working correctly. Instead of affecting only one note/chord, it affects all subsequent ones. In short, currently "fz" is played exactly the same as "mp" and "mf" (the latter are the same in volume). (To convince you: enter a series of identical piano notes, place "mp", "mf" and "fz" everywhere, click play and you will not hear the loudness difference throughout.) It's about MuseScore Basic. In Muse Sounds, everything is almost the same, only "mp" and "mf" differ slightly in loudness, and "fz" is something in between "mp" and "mf".

In reply to by Marc Sabatella

By the way, please note that in Muse Sounds, as well as in MuseScore Basic, "fz" does not affect the sound, which suggests that there is still a relationship with a similar interpretation.
In this regard, I ask once again, can you help me see the list of accent ratios by velocity? + You have not answered the question as to what a "velocity change" is.

Do you still have an unanswered question? Please log in first to post your question.